Source-linked AI summary
Intelligent Omni-Surfaces for Full-Dimensional Wireless Communications: Principle, Technology, and Implementation
Hongliang Zhang, Shuhao Zeng, Boya Di, Yunhua Tan, Marco Di Renzo, Merouane Debbah, Lingyang Song, Zhu Han, H. Vincent Poor
TL;DR
Conventional reflective metasurfaces cover users on only one side, motivating a surface that supports full-dimensional communications. The paper introduces an IOS with joint reflection and refraction, develops hybrid beamforming, and evaluates a prototype experimentally. The experiments substantiate service to users on both sides and show that spectral efficiency depends strongly on IOS configuration.
Problem
Conventional intelligent reflecting surfaces reflect signals toward users on one side, restricting coverage relative to the surface.
Method
The paper introduces an IOS with jointly engineered reflection and refraction, a hybrid beamforming scheme, and an IOS-based wireless communication prototype.
Results
Experimental results substantiate that the IOS can serve users on both sides, while spectral efficiency is highly related to IOS configuration.
Takeaways & Limitations
IOS-based communications can provide full-dimensional coverage, with element states adjusted according to user spatial distribution to achieve maximum system capacity.
Takeaways & Limitations
IOS deployment and beamforming remain constrained by channel-estimation, phase-quantization, coordination, and deployment-optimization challenges.
Abstract
from arXiv · showhide
The recent development of metasurfaces has motivated their potential use for improving the performance of wireless communication networks by manipulating the propagation environment through nearly-passive sub-wavelength scattering elements arranged on a surface. However, most studies of this technology focus on reflective metasurfaces, i.e., the surface reflects the incident signals towards receivers located on the same side of the transmitter, which restricts the coverage to one side of the surface. In this article, we introduce the concept of intelligent omni-surface (IOS), which is able to serve mobile users on both sides of the surface to achieve full-dimensional communications by jointly engineering its reflective and refractive properties. The working principle of the IOS is introduced and a novel hybrid beamforming scheme is proposed for IOS-based wireless communications. Moreover, we present a prototype of IOS-based wireless communications and report experimental results. Furthermore, potential applications of the IOS to wireless communications together with relevant research challenges are discussed.
I. INTRODUCTION
Metasurfaces offer low-cost control of wireless propagation, but conventional intelligent reflecting surfaces cover only users on one side. The proposed intelligent omni-surface (IOS) jointly reflects and refracts signals to support full-dimensional communications, with hybrid beamforming, a prototype, applications, and challenges presented.
- Motivation: Metasurfaces use nearly-passive sub-wavelength elements to adjust signal amplitude and phase, shaping propagation at reduced cost, size, weight, and power.PIN-diode states can configure the scattering response of the surface.
- Motivation: Conventional intelligent reflecting surfaces reflect signals toward users on the same side, leaving users on the opposite side outside coverage.This one-sided operation motivates the IOS concept.
- IOS Concept: The IOS simultaneously reflects and refracts incident signals toward mobile users on both sides, with the reflected-to-refracted power ratio optimized through element structure.Its dual functionality enables full-dimensional communications wherever users are located relative to the surface.
- Contributions: The paper introduces the IOS working principle and proposes hybrid beamforming using digital beamforming at the base station and analog beamforming at the IOS.The scheme is designed to serve mobile users on both sides of the surface.
- Contributions: An IOS-based prototype and experimental evaluation substantiate directional control of reflected and refracted beams for users on both sides.The paper also discusses coverage extension, interference cancellation, secure communications, sensing, localization, and deployment challenges.
II. IOS: WORKING PRINCIPLE
An IOS is an electrically controllable surface whose elements configure amplitude and phase responses. Each incident signal is divided into reflected and refracted components, enabling adaptive coverage on either side.
- IOS Structure: An IOS is a two-dimensional engineered surface comprising electrically controllable scattering elements.The considered surface uses reconfigurable elements with metallic patches and PIN diodes on a dielectric substrate.
- IOS Structure: Each element’s PIN-diode ON/OFF configuration determines the amplitude and phase response applied to incident signals.The elements can be configured across multiple discrete states.
- Signal Operation: An incident signal is fractionally reflected toward the incident side and refracted toward the opposite side of the surface.This behavior applies when signals impinge from either side.
- Signal Operation: The IOS adapts coverage to users on either side by configuring element states and the resulting reflection and refraction responses.The coefficients can depend on incidence direction and the applied phase shift.
III. IOS-BASED HYBRID BEAMFORMING
The IOS-based hybrid beamforming scheme jointly uses base-station digital beamforming and IOS analog beamforming to shape reflected and refracted beams toward users on both sides.
- Hybrid Beamforming: Digital beamforming at the base station and analog beamforming at the IOS jointly shape radio waves toward specified reflection and refraction directions.The base station encodes multiple data streams digitally before transmission through multiple antennas.
- Hybrid Beamforming: The base station allocates transmit power across antennas while the IOS configures the propagation response for users on both sides.The reflected component serves users on the base-station side, while the refracted component serves users on the opposite side.
A. Instantaneous CSI at the BS
Even with full instantaneous channel state information at the base station, optimizing the digital beamformer and IOS configuration remains difficult because their variables are coupled and IOS states are discrete.
- Optimization Challenges: The digital and analog beamformers are coupled, producing a non-convex optimization problem.This coupling complicates joint optimization of the base-station and IOS beamformers.
- Optimization Challenges: Finite IOS element states make the analog-beamformer feasible set discrete, yielding an NP-hard integer program.Sequential optimization with sub-optimal digital beamforming schemes is proposed to reduce computational complexity.
B. Statistical CSI at the BS
Large IOSs create substantial overhead for acquiring instantaneous CSI at the BS, motivating reduced-CSI optimization algorithms. One proposed approach uses slowly varying CSI statistics for IOS analog beamforming while retaining instantaneous CSI for digital beamforming.
- Large IOS element counts create substantial overhead when acquiring instantaneous CSI at the BS.
- Reduced-CSI optimization algorithms are needed to decrease channel-estimation and configuration overhead.
- A proposed solution optimizes the IOS analog beamformer using slowly varying CSI statistics and the digital beamformer using instantaneous CSI.The statistics can include first- and second-order moments obtained from training pilots.
IV. IOS HARDWARE PROTOTYPE
The IOS hardware prototype uses reconfigurable scattering elements controlled through PIN diodes and grouped for simplified operation. Its design supports configurable reflection and refraction responses, with simulated coefficients reported for two operating states.
- The prototype comprises 640 reconfigurable elements, each measuring 2.87 × 1.42 × 0.71 cm3.
- Each IOS element uses metallic patches, a dielectric substrate, and PIN diodes connecting the patches to ground.The PIN diodes control the element response through their ON and OFF states.
- At 3.6 GHz with a 24 MHz transmission bandwidth, Table I reports simulated reflection and refraction coefficients for the two implemented states.
- The IOS is divided into 16 groups of 5 × 8 elements, with each group controlled using the same state.An FPGA automatically changes the group states.
- The FPGA controls IOS group states, while Verilog implements its hardware description language.
B. Hardware Modules of the Prototype
The prototype integrates software-defined transmit and receive chains, synchronization, networking, and centralized data processing. These modules coordinate signal generation, acquisition, processing, and IOS control.
- The transmitter uses a USRP with an RF daughterboard, GNU Radio, an LNA, and a directional double-ridged horn antenna.
- The receiver uses a USRP connected to an LNA and directional double-ridged horn antennas.An external 10MHz OCXO clock provides precise synchronization to the transmitter and receiver.
- A signal synchronizer supplies reference clock and pulse-per-second signals for transmitter and receiver modulation and demodulation.
- An Ethernet switch connects the transmitter, receiver, and host computer for transmitted and received signal acquisition.Its bandwidth is of the order of 1 GHz.
- A Python-controlled host computer coordinates the USRPs and FPGA and processes received signals.The FPGA converts host commands into IOS control signals using Verilog.
V. EXPERIMENTAL EVALUATION
Experiments evaluate IOS beamforming in an isolated room with receivers placed on both sides of the surface. Measurements agree with simulations and demonstrate directional beams and full-dimensional communications, while spectral efficiency depends on IOS configuration.
- The prototype operates in a wave-absorbing room with one transmitter and two receivers positioned on opposite sides of the IOS.The two receivers are placed at the same distance from the relevant setup reference.
- 14 dB and 8 dB are the received SNRs of Rx 1 and Rx 2, respectively.
- Simulations and measurements show good agreement, validating directional beams toward intended users on both IOS sides.
- Two IOS configurations produce two different directional radiation patterns.
- Full-dimensional communications are achieved, while spectral-efficiency distributions depend on IOS-element configuration.The authors state that IOS states should be adjusted according to user spatial distribution to achieve maximum system capacity.
VI. POTENTIAL USE CASES AND RESEARCH CHALLENGES
IOS can extend cellular coverage by serving users on both sides of its surface, but deployment requires joint beamforming, accurate channel acquisition, and location-aware design.
- Coverage Extension: IOS deployment at a cell edge can serve users inside and outside the cell coverage area.This application uses the IOS’s ability to reflect and refract signals toward users on both sides.
- Beamforming Design: Larger coverage requires jointly designing digital beamforming at the base station and analog beamforming at the IOS.Phase-shift quantization creates practical constraints, while balancing performance across both sides remains challenging.
- Channel Acquisition: IOS analog beamforming depends strongly on accurate cascaded-channel CSI, whose estimation overhead increases with the number of IOS elements.CSI from both sides must also be estimated jointly because the channels are correlated.
- IOS Deployment: Optimal IOS deployment remains open because performance depends on its location relative to the transmitter and users.Deployment must account for the trade-off between refractive and reflective functions.
B. Interference Cancellation
In overlapping small-cell coverage areas, an IOS can direct reflected and refracted signals toward intended users while reducing interference toward nonintended users, but coordination remains challenging.
- Interference Motivation: Overlapping coverage among densely deployed small cells can increase multi-cell interference.The IOS is proposed for deployment in these overlapping areas.
- Interference Cancellation: An IOS can reflect and refract signals toward intended cell-edge users while nulling signals toward nonintended users.This configuration is described as an application for alleviating multi-cell interference.
- Research Challenges: Coordination is difficult when multiple small cells share one IOS because each base station lacks CSI for users associated with neighboring cells.The IOS analog beamforming affects users across nearby small cells, motivating efficient coordination protocols.
D. Sensing and Localization
IOS is promising for RF sensing and localization because it can customize propagation channels, enhance signal differences, and reduce out-of-coverage areas. However, IOS-assisted sensing and localization still face optimization and classification challenges.
- D. Sensing and Localization: IOS can customize propagation channels and enhance differences between received signals, supporting accurate RF sensing and localization.Signals received at different locations should be as distinguishable as possible for high accuracy.
- D. Sensing and Localization: The IOS’s full-dimensional communication capability can reduce out-of-coverage areas in sensing and localization applications.
- D. Sensing and Localization: IOS-assisted sensing and localization require analog-beamforming optimization to reduce errors, with compressed sensing and machine learning proposed as possible approaches.Compressed sensing may address sparse signals, while machine learning may classify received signals by objects and user locations.
- D. Sensing and Localization: The paper identifies sensing and localization as an open application and research problem for IOS-assisted wireless communications.